Rosuvastatin, a 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitor used to lower blood low-density lipoprotein cholesterol, is a substrate of the membrane ABCG2 exporter. ABCG2 variants have been shown to alter rosuvastatin disposition. The objective of this study is to determine the impact of ABCG2 34/421 compound haplotypes on rosuvastatin pharmacokinetics in healthy Chinese volunteer subjects. Eight hundred healthy Chinese males were genotyped by polymerase chain reaction–pyrosequencing for ABCG2 34G>A, ABCG2 421C>A, SLCO1B1 521T>C, and CYP2C9*3 variants. Sixty-two male subjects with wild-type SLCO1B1 c.521TT and CYP2C9*3 were recruited for this pharmacokinetic study of rosuvastatin. A single oral dose of 10 mg rosuvastatin was administrated to each subject, and blood samples were collected before and at various time points after drug administration. Plasma concentration of rosuvastatin was determined by high-performance liquid chromatography–tandem mass spectrometry, and pharmacokinetic analysis was carried out using the WinNonlin program. In Chinese males, high allele frequency of ABCG2 c.34G>A (0.275) and c.421C>A (0.282) was observed, resulting in a considerable portion (23.3%) of subjects being ABCG2 34/421 compound heterozygotes. Compared with subjects with ABCG2 wild-type (c.34GG/421CC), plasma rosuvastatin Cmax and area under the curve, AUC0–∞, were significantly higher, while the apparent oral clearance, CL/F, was significantly lower in subjects with c.34AA, c.421AA, and c.34GA/421CA genotypes. Both t1/2 and Tmax were similar among subjects with different genotypes. A high frequency of ABCG2 c.34G>A and c.421C>A variants was present in Chinese males, and the disposition of rosuvastatin was significantly affected by both variants. These data suggest that it is advisable to genotype these variants when prescribing rosuvastatin to Chinese subjects, leading to a precise dose for each individual.
Statins, through reduction of plasma levels of low-density lipoprotein cholesterol (LDL-C), are widely used in the treatment of hyperlipidemia, and subsequently for the prevention of coronary heart disease. Rosuvastatin, a synthetic statin, is a stronger 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitor, and more effective in lowering LDL-C than other statins such as atorvastatin and pravastatin (Jones et al., 2003). However, there is considerable interindividual variability in the efficacy and toxicity of rosuvastatin. Clinical trial data suggest that Asian Americans may have higher plasma drug levels and are at greater risk for rosuvastatin-produced side effects than the general population (U.S. Food and Drug Administration; http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2005/ucm108414.htm). As a result, the Food and Drug Administration has recommended lower rosuvastatin doses in Asians (http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm124906.htm). This variability in rosuvastatin efficacy and toxicity is partially due to genetic variation in the membrane influx and efflux transporters.
ABCG2, also known as the breast cancer resistance protein, is a membrane efflux transporter that is expressed in various normal tissues such as the small intestine, colon, liver, and kidney, and plays a significant role in the disposition of various drugs (Hardwick et al., 2007; Gradhand and Kim, 2008). Currently, more than 50 ABCG2 single-nucleotide polymorphisms (SNPs) have been reported in various ethnic populations (Iida et al., 2002; Bäckström et al., 2003; Zamber et al., 2003; Kobayashi et al., 2005). Previous studies have shown a significantly higher frequency of ABCG2 c.34G>A (rs2231137) and c.421C>A (rs2231142) variants in Asians when compared with Caucasians and African Americans (Kim et al., 2010). The ABCG2 c.421C>A polymorphism has been reported to markedly affect the pharmacokinetics, efficacy, and toxicity of rosuvastatin (Zhang et al., 2006; Keskitalo et al., 2009b). However, the importance of ABCG2 c.34G>A polymorphism in rosuvastatin disposition has not been studied in humans. Thus, the objective of this study was to determine the influence of the ABCG2 c.34G>A SNP in the rosuvastatin pharmacokinetics in healthy Chinese volunteers. Because ABCG2 c.34G>A and c.421C>A variants are not in linkage disequilibrium (Kwan et al., 2011) and the reports of c.421CA heterozygotes on rosuvastatin disposition are inconsistent (Zhang et al., 2006; Keskitalo et al., 2009b), the effect of the ABCG2 34GA/421CA compound heterozygotes on the rosuvastatin pharmacokinetics was also analyzed.
Materials and Methods
Genomic DNA Extraction and Determination of SLCO1B1 c.521T>C, ABCG2 c.421C>A, and ABCG2 c.34G>A Polymorphism.
Genomic DNA from 800 healthy Chinese men was extracted from whole blood samples using the QuickGene DNA whole blood kit (Kurabo, Osaka, Japan), and the DNA concentration was determined using a Nanodrop 2000 (Thermo Scientific, Wilmington, DE). In order to determine the c.521T>C polymorphism of the SLCO1B1 gene, a 119 bp fragment was amplified by polymerase chain reaction (PCR) using a specific pair of primers (Table 1) under the following conditions: one cycle predenaturation at 95°C for 5 minutes, 35 cycles of denaturation at 95°C for 30 seconds, annealing at 53°C for 30 seconds, and elongation at 72°C for 30 seconds; and one cycle of postelongation at 72°C for 5 minutes. The ABCG2 c.34G>A and c.421C>A polymorphic regions were amplified by PCR using the primer pairs indicated in Table 1 under the same conditions as SLCO1B1. The single strand sequencing templates were purified using the Pyromark ID instrument (Qiagen, Hilden, Germany). The SNP was determined by pyrosequencing on a Pyromark Q96 ID platform (Qiagen) using PyroMark Gold Q96 Reagents (Qiagen) with the specific primers indicated in Table 1 and following the manufacturer’s instructions. The resulting polymorphisms were identified automatically by the PyroMark ID software (Qiagen), and verified by manual analysis. The CYP2C9*3 polymorphism was determined by PCR pyrosequencing using specific primers (Table 1) as previously described (Zhao et al., 2009). Three to five samples for each polymorphism were randomly selected and directly sequenced using the Sanger chain-termination method in order to verify the accuracy of the pyrosequencing method, and the results were completely in agreement with the pyrosequencing (Wan et al., 2012a,b).
Subjects and Pharmacokinetic Study Design.
A total of 62 healthy Chinese men of Han ethnicity, between ages 18 and 24, were recruited for the pharmacokinetic study. These subjects had a body mass index ranging from 18 to 24 and they were considered healthy as ascertained by physical examination. All subjects were nonsmokers and free of drugs, and had abstained from coffee, tea, and alcohol for at least 1 week before participating in the study.
These subjects were divided into six groups according to the ABCG2 c.[34G>A(;)421C>A] haplotypes as indicated in Tables 2 and 3. All subjects had a SLCO1B1 c.521TT genotype and a CYP2C9 wild type (WT). After an overnight fast, they received a single oral dose of a 10 mg rosuvastatin calcium tablet (AstraZeneca, Wuxi, Jiangsu, China) taken with 200 ml of water. Meals were allowed 6 hours later after rosuvastatin administration. Venous blood was collected before rosuvastatin administration and at 0.5-, 1-, 1.5-, 2-, 3-, 4-, 5-, 6-, 8-, 10-, 12-, 24-, 36-, and 48-hour intervals afterward. The plasma was separated by centrifugation and stored in polypropylene tubes at −40°C until analysis.
This study was approved by the Ethics Committee of Xiangya Medical College, Central South University (Changsha, China), and registered in the Chinese Clinical Trial Registry (ChiCTR-RCH-12002706; http://www.chictr.org/en/). Written informed consent was obtained from each participant. This study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments (World Medical Association Declaration of Helsinki, 2001).
Determination of Plasma Rosuvastatin Concentration and Pharmacokinetic Analysis.
The plasma concentration of rosuvastatin was determined using the API4000 high-performance liquid chromatography–tandem mass spectrometry system with an electrospray ionization source system and Analyte software, version 1.4 (Applied Biosystems, Foster, CA). A Hypurity C18 column (Thermo Fisher Scientific Inc., Waltham, MA) and a mobile phase, consisting of 2 mM ammonium acetate in 0.2% formic acid and acetonitrile (1:1, v/v) at a flow rate of 0.3 ml/min, were applied. A plasma sample of 450 μl was mixed with 50 μl of 57.78 ng/ml atorvastatin (Toronto Research Chemicals, North York, ON, Canada), 50 μl of 10% acetic acid, and 2 ml of diethyl ether, vortexed for 10 minutes, and then centrifuged at 4000 rpm for 10 minutes. The supernatant was removed and evaporated to dryness at 42°C under a gentle stream of nitrogen. The residue was reconstituted in 150 μl of the mobile phase, and 10 μl of the sample was injected into the liquid chromatography–tandem mass spectrometry system for analysis. The mass spectrometer was operated in an electrospray ionization positive-ion mode. The multiple reaction monitoring transitions were performed at m/z 482.1–258.1 for rosuvastatin and m/z 559.4–440.3 for atovastatin. The limit of quantification was 0.369 ng/ml for rosuvastatin and the linear range of the assay was 0.369–88.5 ng/ml. The interday accuracies for rosuvastatin were within 85–115% and the total imprecision was <5%.
Pharmacokinetic analysis of rosuvastatin was performed using the WinNonlin program, version 4.1 (Pharsight Corporation, Mountain View, CA). The Cmax and Tmax values were determined directly from the plasma concentration/time data, and the half-life (t1/2) was calculated from the slope obtained by log-linear regression of the terminal plasma concentration/time data. The area under the curve (AUC) was determined by the linear trapezoidal method. The apparent oral clearance (CL/F) was calculated as follows: CL/F = dose (10 mg)/AUC(0–∞) (ng·h/ml).
Pharmacokinetic data are expressed as mean ± S.E.M. in Figs. 1–3 and as mean ± S.D. in Table 3. Genetic equilibrium was tested according to the Hardy-Weinberg formula using the χ2 test. Statistical analyses of the pharmacokinetic data were carried out using SPSS version 14.0 (SPSS Inc., Chicago, IL). Differences among multiple groups were analyzed by one-way analysis of variance followed by the post hoc Student-Newman-Keuls test. A P value of less than 0.05 was considered statistically significant.
Frequency of ABCG2 c.421C>A and c.34G>A Polymorphisms in the Chinese Population.
The ABCG2 c.421C>A and c.34G>A polymorphisms were simultaneously determined in 800 healthy Chinese men by PCR pyrosequencing. As shown in Table 2, the allele frequencies of c.34G>A and c.421C>A were 0.275 and 0.282 in this population, respectively. The allele frequency of c.421C>A in this population met the Hardy-Weinberg equilibrium (χ2 = 0.714, P = 0.398), while c.34G>A was not in Hardy-Weinberg equilibrium (χ2 = 51.58, P < 0.001). The 34GG and 34AA homozygotes at c.34G>A accounted for 57.6 and 12.6% of allele frequency, and 421CC and 421AA at c.421C>A accounted for 50.9 and 7.4% of allele frequency, respectively. Six haplotypes were detected and their frequencies are presented in Table 2. Subjects with the c.[34GG;421CC] haplotype, which is considered the reference WT, accounted for 31.7% of this population. The c.[34AA;421CC] and c.[34GG;421AA] haplotypes were 12.6 and 7.4%, respectively. Interestingly, a very high percentage (23.3%) of compound heterozygotes of c.[34GA;421CA] variants was determined in this population. The frequency of the SLCO1B1 c.521T>C SNP was also determined in this population, and the frequencies of the 521TT, 521TC, and 521CC genotypes were 80, 19, and 1%, respectively, which was in Hardy-Weinberg equilibrium (χ2 = 0.095; P = 0.758).
Effect of ABCG2 421C>A and 34G>A Polymorphisms on Rosuvastatin Pharmacokinetics.
Sixty-two subjects selected from 800 healthy Chinese volunteers participated in the rosuvastatin pharmacokinetics study. All 62 subjects had a SLCO1B1 c.521TT genotype and a CYP2C9 WT. Of the 62 individuals, 10 subjects had a haplotype of c.[34GG;421CC], nine had a haplotype of c.[34GA;421CC], nine had a haplotype of c.[34AA;421CC], eight had a haplotype of c.[34GG;421CA], 10 had a haplotype of c.[34GG;421AA], and 16 had a haplotype of c.[34GA;421CA].
Considering the c.421C>A SNP in subjects with an ABCG2 c.34GG homozygotes, subjects with c.421AA homozygotes had 227 and 200% higher mean Cmax values of rosuvastatin than those with c.421CC and c.421CA genotypes (P < 0.001), as shown in Fig. 1 and Table 3. The mean AUC(0–∞) values of rosuvastatin were 164 and 154% greater and the mean AUC(0–48 h) values were 164 and 155% greater in the c.421AA homozygotes than in those with the c.421CC and c.421CA genotypes (P < 0.001). In contrast, the mean CL/F values in c.421AA homozygotes were 63 and 64% lower (P < 0.001) than in subjects with c.421CC and c.421CA genotypes (Table 3). However, there were no statistically significant differences in AUC(0–∞), AUC(0–48 h), and Cmax between c.421CC homozygotes and c.421CA heterozygotes or in t1/2 and Tmax of rosuvastatin among the six studied haplotypes (Table 3).
Similar to the c.421C>A SNP, the c.34G>A SNP in the c.421CC homozygous background also had a significant influence on the rosuvastatin pharmacokinetics as shown in Fig. 2 and Table 3. Subjects with the c.34AA genotype had significantly higher Cmax, AUC(0–∞) and AUC(0–48 h) values with a lower CL/F value than those with the c.34GG and c.34GA genotypes (P < 0.01). However, the alteration in rosuvastatin pharmacokinetics in subjects with the c.34AA homozygote was significantly less compared with those with c.421AA homozygotes (see Fig. 3; Table 3). The AUC0–∞, AUC0–48 h, and Cmax values of rosuvastatin were 64, 64, and 57% greater, while the CL/F value was 44% lower in subjects with c.421AA homozygotes than in those with c.34AA homozygotes (P < 0.05). The Cmax, AUC(0–∞), and AUC(0–48 h) values were not statistically different between c.34GG homozygotes and c.34GA heterozygotes.
Although neither c.421CA nor c.34GA heterozygotes had a significant influence on the rosuvastatin pharmacokinetics, the AUC0–∞, AUC0–48 h, and Cmax values of rosuvastatin in the compound c.[34GA(;)421CA] heterozygotes were 69, 67, and 74% higher, while the CL/F value was 40% lower than those with the c.[34GG;421CC] WT reference group (P < 0.01) (Fig. 3; Table 3). However, there were no statistically significant differences in t1/2 and Tmax between these two groups.
Our present study has analyzed the frequency of ABCG2 c.34G>A and c.421C>A variants and their influence on the rosuvastatin pharmacokinetics in a large population of healthy Chinese males. For the first time, we have demonstrated that both ABCG2 c.34G>A and c.421C>A variants play a significant role in the rosuvastatin pharmacokinetics in humans. In agreement with previous studies (Zhang et al., 2006; Keskitalo et al., 2009b; Lee et al., 2013), the ABCG2 c.421C>A variant has a significant influence on the rosuvastatin pharmacokinetics. The plasma Cmax and AUC values of rosuvastatin in c.421AA homozygotes were about 200% greater than those with c.421CC homozygotes (see Table 3), which is similar to results reported in previous studies (Zhang et al., 2006; Keskitalo et al., 2009b; Lee et al., 2013). Moreover, in c.34AA homozygotes, the plasma Cmax and AUC values of rosuvastatin were significantly higher than those with c.34GG homozygotes, indicating that the c.34G>A variant also plays a significant role in the rosuvastatin pharmacokinetics. Recently, Kim et al. (2015) reported that the ABCG2 c.421C>A variant significantly influenced the pharmacokinetics of bicalutamide in a gene dose-dependent manner but that c.34G>A did not have such an effect on healthy Korean male subjects. It is unclear whether the lack of change in the bicalutamide pharmacokinetics in subjects with the c.34AA genotype was due to the small sample size of the study or a drug-specific action of the ABCG2 genetic variants. The reason remains to be elucidated.
Most importantly, we demonstrated in the present study that the ABCG2 c.34G>A and c.421C>A compound heterozygotes—but not the heterozygotes of either c.34GA or c.421CA—had a significant impact on rosuvastatin disposition, resulting in elevated plasma Cmax and AUC values (see Table 3). This demonstration in compound heterozygotes may explain, at least in part, the inconsistent reports on the rosuvastatin pharmacokinetics in c.421CA heterozygotes. Zhang et al. (2006) and Lee et al. (2013) reported significant differences in the plasma Cmax and AUC values of rosuvastatin in c.421CA heterozygotes when compared with c.421CC homozygotes of healthy Chinese males. These changes were not observed in c.421CA heterozygous Caucasians (Keskitalo et al., 2009b). This discrepancy is most likely due to the high frequency of compound heterozygotes in Chinese populations, accounting for over 55% of c.421CA heterozygotes (see Table 2). In contrast, both the c.34A and c.421A alleles are quite low in Caucasian populations (Kim et al., 2010). Taken together, these data indicate that genetic variants of the ABCG2 efflux transporter play a significant role in the disposition of rosuvastatin in humans.
Changes in the rosuvastatin pharmacokinetics could subsequently result in a significant impact on rosuvastatin efficacy and toxicity. Tomlinson et al. (2010) found that in Chinese patients with hypercholesterolemia, subjects with the c.421AA variant had a significantly greater reduction in LDL-C level than those with the c.421CC variant after a daily 10 mg rosuvastatin regimen, and this reduction was equivalent to the effect obtained by doubling the dose of rosuvastatin. Lee et al. (2013) reported that, in Chinese patients with hypercholesterolemia, subjects with the ABCG2 c.421AA genotype had a higher plasma concentration of rosuvastatin and a greater reduction in LDL-C than those with the c.421CC genotype when they were treated with the same rosuvastatin regimen. Furthermore, the Food and Drug Administration has warned of a higher incidence of rosuvastatin toxicity in Asian subjects compared with Caucasians when they used the same rosuvastatin regimen (http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2005/ucm108414.htm). Therefore, it is imperative to personalize rosuvastatin treatment in patients, especially in Asians, because of the high frequency of ABCG2 c.34A and c.421A alleles.
The elevated plasma concentration of rosuvastatin in c.421AA homozygotes is consistent with previous demonstrations of a lower expression level of the ABCG2 efflux transporter protein and a reduced ability to export substrate for the c.421AA variant. This leads to increased drug absorption in the gastrointestinal track, and drug accumulation in hepatocytes and systemic circulation (Kondo et al., 2004; Morisaki et al., 2005; Urquhart et al., 2008). On the other hand, the ABCG2 c.34AA variant has been shown to possess decreased efflux activity due to disturbed protein membrane localization, which could also result in increased drug absorption in the gastrointestinal track and drug accumulation in systemic circulation (Mizuarai et al., 2004). However, the molecular mechanism of how the compound c.34GA/421CA heterozygote affects the rosuvastatin pharmacokinetics remains to be investigated.
The disposition of rosuvastatin may also be influenced by other genetic variations in addition to ABCG2 variants. It has been shown that the SLCO1B1 c.521T>C variant was able to alter rosuvastatin disposition in humans (Bolego et al., 2002; Pasanen et al., 2007). To minimize the influence of SLCO1B1 c.521T>C in the rosuvastatin pharmacokinetics, we genotyped the SLCO1B1 c.521T>C variant in subjects, and only those subjects with the SLCO1B1 c.521TT genotype were included in the study. Furthermore, because approximately 10% of rosuvastatin is metabolized by CYP2C9 (Bolego et al., 2002), we genotyped CYP2C9*3 in the volunteers and confirmed all 62 participants possessed the CYP2C9 WT genotype (unpublished data). Although it has been shown that rosuvastatin is a substrate of ABCB1 and ABCC2 transporters, pharmacokinetic analysis indicates that the genetic variations of ABCB1 and ABCC2 had no significant impact on rosuvastatin disposition in humans (Kitamura et al., 2008; Keskitalo et al., 2009a). Taken together, these data suggest that the observed variation in rosuvastatin disposition is mainly due to genetic variants of the ABCG2 gene.
Finally, to the best of our knowledge, the present study is the largest study of the frequency of ABCG2 c.34G>A and c.421C>A variants in healthy Asian males with a sample size of 800. In agreement with previous studies (Kim et al., 2010), we confirmed that the Chinese population, as in other Asian populations, has the highest incidence of ABCG2 c.34G>A and c.421C>A variants with allele A frequencies of 0.275 and 0.282, respectively (see Table 2), compared with frequencies of 0.175 and 0.103 in Caucasian and 0.10 and 0.023 in African-American populations (Kim et al., 2010). It is obvious that this high allele frequency is an important factor that should be considered in the administration of rosuvastatin in Asian populations. Additionally, a deviation of the Hardy-Weinberg equilibrium for the c.34G>A, but not the c.421C>A, variant was observed in this population. Although disequilibrium may arise from inbreeding, we were not aware of any consanguinity within the population as we recruited the subjects from a large urban area. As previously suggested, a deviation could be a consequence of the recruiting strategy despite the fact that the other tested polymorphism did not deviate from the Hardy-Weinberg equilibrium (Böger et al., 2005, 2007; Kim et al., 2010). A potential explanation is a sampling bias because all subjects were male. Further study is necessary to confirm this hypothesis.
In conclusion, we have demonstrated that both ABCG2 c.34G>A and c.421C>A variants play a significant role in the rosuvastatin pharmacokinetics, almost doubling the Cmax and AUC values of rosuvastatin in c.34AA and c.421AA homozygotes and in c.34GA/c.421CA compound heterozygotes compared with the reference WT group. Considering the high allele A frequencies of c.34G>A and c.421C>A variants in Chinese and other Asian populations, it is advisable to personalize the dosage of rosuvastatin in the treatment of hypercholesterolemia through genotyping genetic variations, such as ABCG2, in patients.
The authors thank Jeff Zhu for carefully revising this manuscript.
Participated in research design: Wang, Zhu.
Conducted experiments: Wan, Wang, Li, Xu, Pei, Peng, Sun, Cheng, Zeng.
Contributed new reagents or analytic tools: Pei, Yang.
Performed data analysis: Wan, Wang, Zhu.
Wrote or contributed to the writing of the manuscript: Wan, Wang, Zhu.
- Received April 9, 2015.
- Accepted June 12, 2015.
Z.W. and G.W. contributed equally to this work.
This work was supported in part by the grants from the National Natural Scientific Foundation of China [Grants 81072706, 81302851, and 81403021]; the Science and Technology Project of Hunan Province, China [Grants 2013FJ3036 and 2014FJ3023]; and a Special Talents Fund from Central South University of China.
- area under the curve
- low-density lipoprotein cholesterol
- polymerase chain reaction
- single-nucleotide polymorphism
- wild type
- Copyright © 2015 by The American Society for Pharmacology and Experimental Therapeutics